AU2013263818A1

AU2013263818A1 - Method for detecting locomotive traction motor speed signals using magnetoelectric sensors

Info

Publication number: AU2013263818A1
Application number: AU2013263818A
Authority: AU
Inventors: Fuzhi CAO; Yabin DU; Jinglai LIU; Ruizhen ZHANG
Original assignee: CNR Dalian Locomotive and Rolling Stock Co Ltd
Current assignee: CRRC Dalian Co Ltd
Priority date: 2012-12-13
Filing date: 2013-03-08
Publication date: 2014-07-03
Anticipated expiration: 2033-03-08
Also published as: CN103105503A; CN103105503B; NZ618327A; WO2014089938A1; AU2013263818B2

Abstract

The invention provides a method which comprises recording triggering time of present and previous rising or falling edges of sensor pulse signals, obtaining present time difference ATm, frequency Fm= 1/ATm, and a present detection value of traction motor speed. When the locomotive operates in a traction or braking mode, current value Im of the traction motor is detected, and current average value Iavr, current imbalance degree Fim, and speed average value Navr of the traction motor is calculated. If Nm <Navr x f and P-m <Fmax continuously hold true for more than a preset number of times, R, it may determine that an m-th sensor is in a fault state. When the locomotive is in a non-traction and non-braking mode, if Nm <Navr x r holds true more than a preset number of times, R, it may determine that the m-th sensor is in a fault state. For a non-fault m-th sensor, if Nm> Nmc + A N is true, let Nnc = Nmc + A N; otherwise, if Nm <Nmc - A Nm holds true, let Nmc = Nme - A N; otherwise, let Nme = Nm. A fault sensor can be removed, interference can be avoided and the real-time requirements for speed feedback when wheel slip or skid happens can be met.

Description

Method for detecting locomotive traction motor speed signals using magnetoelectric sensors FIELD OF THE TECHNOLOGY [0001] The invention relates to railway locomotive technology, in particular to a method for detecting locomotive traction motor speed signals using magnetoelectric sensors. BACKGROUND [0002] When locomotives moves on rails, traction force and braking force are highly subject to the adhesion between wheels and the rail. With the continuous development of modern railway, the load of freight trains and the speed of passenger trains constantly increase. Fully taking advantage of wheel-rail adhesion and effectively preventing wheel slip in traction or wheel skid in braking has become the direction of development in the railway locomotive braking field worldwide. [0003] Continuous wheel slip or wheel skid will cause wheel tread abrasions. If wheel slip occurs in positioning mode, severe rail pits will be generated by friction. In this situation, the rails need to be replaced for maintenance. It will interfere severely with the normal operation of the railway. If the braking force exceeds the adhesive force, the rotation speed of wheels will be drastically reduced or wheels stops rotating, leading to skid. Wheel skid of the train in braking will cause various problems, such as, wheel and rail becoming heat, abrasion of wheel and rail, and so on. These problems, when becoming serious enough, even will severely affect the safe operation of locomotives and are extremely harmful. [0004] In order to equip trains with efficient, safe and reliable anti-wheel slip and skid control system, a lot of research has been done for a long time. For example, a mathematical model may be built by calculating a speed difference, axis rotation acceleration and differential of the axis rotation acceleration, with respect to each wheel axis, based on the actual measurement of the traction motor speed, and control can be performed by using these parameters. [0005] The rotation speed of a locomotive traction motor is typically detected through the use of speed measurement gears mounted on traction motor pinions, a gap exists between the speed measurement gears and the speed sensors. When wheels rotate, speed sensor will generate alternating frequency signals proportional to wheel speed. Traction motor shafts are coupled to 1 the wheel via pinions and gears. Sensors outputs a certain amount of pulses per cycle of rotation of wheels, so the frequency of the induced alternating signals is proportional to the rotation speed of the locomotive traction motor. For such pulse signals generated by sensors, timing sampling is usually implemented. The amount of pulses within a period of time is detected and then calculation is carried out. The speed signals are often mixed with some high-frequency signals. Due to the existence of these high-frequency signals, incorrect speed value is possibly detected by the control device. Therefore, a digital low-pass filter is included in the signal process means to filter high-frequency noise from the useful signals. Accuracy and real -time performance of this speed measurement method are poor. The control system can only detect the amount of the rising edges of the pulse signals, only accurate to bits. The amount for the teeth of speed measurement gears is very low. Thereby, if the period for the timing sampling is short, although real-time performance can be improved, the accuracy of sampling is poor. If the period for the timing sampling is long, although the accuracy can be improved, the real-time performance will be degraded. When the wheelset of locomotive is wheel slip or skid, as a result of line conditions, vibration, electromagnetic interference and sensors fixation as well as other factors, the pulse signals generated by sensors will be mixed with the high frequency signals, and the pulse signal waveform itself will also be degraded,. the amplitude and the zero crossing point of the pulse signals will change correspondingly, and even some pulses are lost. Another approach relates to performing several detections and averaging the obtained speed signals. Alternatively, the maximum and minimum values of the speed signals are omitted before averaging is carried out. Though the interference signals could be filtered out to some extent, there is a hysteresis in speed signal detection, and the real-time performance of wheel slip and skid control is poor. SUMMARY [0006] An object of the invention is to solve one or more of the above problems, by providing a method for detecting locomotive traction motor speed signals using magnetoelectric sensors. The method can be implemented in an anti-wheel slip and skid control system. [0007] According to an aspect of the invention, a method for detecting locomotive traction motor speed signals using magnetoelectric sensors includes: A. recording triggering time of present and previous rising or falling edges of sensor pulse signals using a timer, calculating present time difference ATm and frequency Fm= 1/ATm, and 2 obtaining a present detection value of traction motor speed, Nm= 60x Fm / Z (Where: Z is the amount of teeth of a speed measurement gear of the traction motor, m is the number of the shaft of the traction motor); B. when the locomotive operates in a traction or braking mode, detecting a current value Im of the traction motor, calculating a current average value Iavr of the traction motor, calculating current imbalance degree em of the traction motor, Fm=| (Im-Iavr) /Iavrl, calculating speed average value Navr of the traction motor, and if Nm <Navr x f and gm <Fmax continuously hold true for more than a preset number of times, R, determining that an m-th sensor is in a fault state; when the locomotive is in a non-traction and non-braking mode, if Nm <Navr x rj hold true more than a preset number of times, R, determining that the m-th sensor is in a fault state,(where, smax is a maximum limit value of current imbalance degree, f is the limit value of speed imbalance degree, and R is the number of times of repetitions); C. if the m-th sensor is non-fault, determining if Nm> Nnc + A N holds true, and if it is true, let Nmc = Nic + A N; otherwise, determining if Nm <Nnc - A Nm holds true, and if it is true, let Nnc = Nme - A N; otherwise, letting Nnc = Nm (where, Nm is speed detection value, Nme is a speed correction value, and AN is a maximum limit value of speed change). [0008] Preferably, the number of the times described in the step B is set in the range 3 to 5. [0009] Preferably, the maximum limit value of speed change described in the step C is obtained according to the following empirical formula : AN = 3600/ (ZxU) (where, Z- is the amount of teeth of the speed measurement gear of the traction motor, and U is a gear ratio). [0010] According to the invention, a fault sensor can be accurately removed. For any non-fault sensor, the speed mutation due to interference, vibration and so on can be avoided, and meanwhile the real-time requirements for speed feedback when wheel slip or skid happens can be met. The affects due to line conditions, vibration, electromagnetic interference, the sensor fixing condition and other factors can be eliminated, and the reliability of control for preventing wheel slip in traction or wheel skid in braking can be improved. BRIEF DESCRIPTION OF THE DRAWINGS [0011] Figure 1 is a flow chart of an input capture interrupt subroutine according to an embodiment of the invention; [0012] Figure 2 is a flow chart of a speed measurement subroutine according to an embodiment 3 of the invention; [0013] Figure 3 is a flow chart of a speed sensor fault diagnosis subroutine according to an embodiment of the invention; [0014] Figure 4 is a flow chart of a current imbalance maximum limit value calculation subroutine according to an embodiment of the invention. DETAILED DESCRIPTION [0015] In order to make the object, technical solutions and advantages of the invention clearer, the following will clearly and comprehensively describe the technical solutions of the embodiments of the present invention with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments , but not all embodiments of the present invention. Any other embodiments obtained by the skilled in the field without creative efforts, based on the teaching of the present invention, fall into the protection scope of the present invention. [0016] Referring to Figure 1, the input capture interrupt subroutine first closes the input capture interrupt in block 1.1 and the input capture interrupt flag is cleared in block 1.2. Then, a traction motor speed measurement subroutine is executed in block 1.3 and input capture interrupt is opened in block 1.4 to be ready for the next input capture interrupt. The interrupt subroutine is ended. [0017] Referring to Figure 2, the traction motor speed measurement subroutine determines if the sensor fault flag is true in block 2.1. If it is true, the process forwards to block 2.13 to let Nmc = 0, otherwise forwards to block 2.2 to read the register TMR2 and let Tint = TMR2. The time difference A T = Tmt-Tml is calculated in block 2.3. Tml = Tit is set in block 2.4. The frequency of the input signal, Fm= 1/AT, is calculated in block 2.5. The traction motor speed, Nm = 60 x Fm / Z, is calculated in block 2.6, where Z is the amount of teeth of the speed measurement gear of the traction motor and m is the number of the shaft of the traction motor. The traction motor speed sensor fault diagnosis subroutine is executed in block 2.7. Determination if Nm> Nmc + ,N is true is made in block 2.8. If it is true, Nmc = Nmc + A N is set in block 2.9, otherwise the process forwards to block 2.10. Determination if Nm< Nmc-'N is true is made in block 2.10. If it is true, the process forwards to block 2.11, letting Nmc = Nmc A N; otherwise, the process forwards to block 2.12, letting Nmc =Nm. The value of A N may be obtained by the empirical formula: ,N = 3600/ (ZxU) , where Z is the number of teeth of the 4 speed measurement gear of the traction motor, and U is a gear ratio. An example for the values of these parameters is Z = 60, U = 93/17, A N = 11. [0018] Referring to Figure 3. the sensor fault diagnosis subroutine determines whether the locomotive operation mode is a traction mode in block 2.8.1. If so, the process forwards to block 2.8.3, otherwise, forwards to block 2.8.2. In block 2.8.2, it determines whether the locomotive operation mode is a braking mode. If so, the process forwards to block 2.8.3, otherwise, to block 2.8.13. In block 2.8.3, current value Im is detected for each traction motor. In block 2.8.4, traction motor current average Iavr is calculated. In block 2.8.5, current imbalance Fm = | (Im Iavr) / Iavr I is calculated for each traction motor. In block 2.8.6, current imbalance maximum limit value smax is calculated. In block 2.8.7, traction motor speed average Navr is calculated. Then the process forwards to block 2.8.8 to determine whether Nm <Navr x f( where f is a limit value for the speed imbalance, and preferably f is 50%) and Fm <Fmax. If so, the process forwards to block 2.8.9, otherwise, to block 2.8.10 where the counter T is set to 0, and the subroutine is ended. In block 2.8.9, the counter T is increased by 1. The process forwards to block 2.8.11, to determine whether T> A T (ranging from 3 to 5). If so, the process forwards to block 2.8.12, where the sensor fault flag is set to true, corresponding alarm message is sent, and the speed signal of this shaft is disregarded so that Nm is not used to determine wheel slip or skid, and the subroutine is ended; otherwise the subroutine is ended directly. The average Navr of traction motor speed N1 N6 is calculated in block 2.8.13. In block 2.8.14, determination whether Nm <Navr x f is made. If so, the process forwards to block 2.8.15, setting T to 0; otherwise the process forwards to block 2.8.16, increasing the counter T by 1.Then the process forwards to block 2.8.17, determining whether T'>AT. If so, the process forwards to block 2.8.18, setting the sensor fault flag to true, sending corresponding alarm message, and disregarding the speed signal of this shaft so that Nm is not used to determine wheel slip or skid, and the subroutine is ended; otherwise the subroutine is ended directly. [0019] Referring to Figure 4, the maximum limit smax of current imbalance is calculated. In block 3.1, determination whether Iavr > Imax * 80% is made. If so, the process forwards to block 3.4 to set smax = 10%,then the subroutine is ended; otherwise determination whether Iavr > Imax* 50% is made in block 3.2. If so, the process forwards to block 3.5 to set smax = 12%, and the subroutine is ended; otherwise the process forwards to block 3.3 to set smax= 15%, and the subroutine is ended. [0020] While the present invention has been shown and described with respect to the preferred 5 embodiments, the present invention is not necessarily limited thereto. It will be easily appreciated by those skilled in the art that various substitutions, changes and modifications may be made within a scope without departing from the technical idea of the invention. 6

Claims

1. A method for detecting a rotation speed signal of a traction motor of a locomotive using a magnetoelectric sensor, comprising: A. recording triggering time of present and previous rising or falling edges of pulse signals of the sensor using a timer, and calculating present time difference ATm and frequency Fm= 1/ATm to obtain a present detection value Nm of the rotation speed of the traction motor, Nm 60x Fm / Z, where Z is the number of teeth of a speed measurement gear of the traction motor, and m is the serial number of a shaft of the traction motor; B. when the locomotive operates in a traction or braking mode, detecting current values Im of the traction motors, calculating an average value Iavr of the current values, calculating an imbalance degree Fm of the current values, where Fm=| (Im-Iavr) /Iavrl, and calculating an average value Navr of the rotation speed, and if Nm <Navr x f and gm <Fmax continuously holds true for more than a preset number of times, R, it is determined that an m-th sensor is in a fault state; when the locomotive is not in the traction mode or the braking mode, if Nm <Navr x ' holds true for more than the preset number of times, R, it is determined that the m-th sensor is in the fault state, where Fmax is a maximum limit value of the imbalance degree of the current values,1 is a limit value of an imbalance degree of the rotation speed, and R is the number of repeated times; C. if the m-th sensor is not in the fault state, determining whether Nm> Nie + A N holds true, and if it is true, let Nmc = Nnc + A N; otherwise, determining whether Nm <Nmc - A Nm holds true, and if it is true, let Nmc = Ninc - A N; otherwise, letting Nmc = Nm, where, Nm is the detection value of the rotation speed, Nmc is a corrected value of the rotation speed, and AN is a maximum limit value of change in increased or reduced rotation speed.

2, The method according to claim 1, wherein the preset number of times in step B is set in the range 3 to 5.

3, The method according to claim 1, wherein the maximum limit value of the change in the increased or reduced rotation speed in step C is obtained according to an empirical formula AN = 3600/ (ZxU) , where, Z is the number of the teeth of the speed measurement gear of the traction motor, and U is a gear ratio. 7